Table Of Content

GROWTH & Hormones Genetics Vol. 14 No. 1 April 1998 The Spectrum of Uses for Grov^th Hormone in Children Ron G. Rosenfeld, MD Letter From the Editor Chairman The two lead articles in this issue are published primarily DepartmentofPediatrics for nonendocrinologists who read GROWTH, Genetics, & Oregon IHealth Sciences University Hormones, although endocrinologists also will undojbt- Portland, Oregon edly appreciate the authors' opinions. The purpose of these articles is stated in the opening sentence of Dr. INTRODUCTION Rosenfeld in hisarticleentitled "The Spectrum of Usesfor Growth Hormone in Children." Specifically, Dr. Rosenfeld This is a reasonable time for a reassessment of the states: "This is a reasonable time for a reassessment of use of growth hormone (GH) in childhood, as pituitary- the use of growth hormone (GH) in childhood." Dr. derived human GH (hGH) was first tested in children Rosenfeld is eminently qualified as an expert both in 40 years ago and recombinant DNA-derived hGH has studying the clinical and laboratoryaspects ofthe GH axis been available for more than a dozen years. Multiple and in caring for patients with disturbances of this axis. controlled and uncontrolled clinical trials have been Dr. Siyper also qualifies as a significant contributor, as is performed and tens of thousands of children world- evident from histhoughtful and cautious approach to the wide are receiving therapy. GH has received Food use of GH as a pharmacologic agent. He published his and Drug Administration (FDA) approval in the United thoughts initially in 1995 in IVIedicalHypothesis'\n an arti- GH cle entitled "How Safe and Effective Is hGH at Pharmaco- States for 3 pediatric indications: childhood defi- logic Dosing?" He has updated his presentation for publi- ciency (GHD), growth failure associated with chronic cation inthis issue of GGH. renal insufficiency (CRI); and short stature associated Other recent important articles concerning the use of with Turner syndrome (TS). Treatment also has been human GH (hGH)thatourreaders maywishto consult in- approved for AIDS-associated wasting and adult GHD. The ability to produce recombinant GH in es- cCloumdmei(t1t)eaen aornticBlieoferthoimctsheofCtohmemiAtmteereicoannDrAucgasdaenmdythoef sentially unlimited quantities has allowed higherdoses Pediatrics, published in Pecf/afr/cs (1997;99:122-129), to be employed for both conventional and less con- that is entitled "Considerations Relatedtothe Use of rhGH GH ventional uses. Provocative testing, for years the in Children"; (2) an article edited by Cuttler et al, entitled definitive diagnostic method forGHD, has come under "Short Stature and GH Therapy: A National Study of renewed criticism; alternative diagnostic strategies, Physician Recommendation Patterns," published in such as 24-hour GH sampling, quantitative excretion JAMA (1996;276:531-537); and (3) an article from a of urinary GH, and determinations of serum concen- committee appointed from the Lawson Wilkins Pediatric trations of insulin-like growth factor 1 (IGF-1), IGF- Endocrine Society, entitled "Guidelines for the Use of binding protein 3 (IGFBP-3) and, possibly, the acid la- GH in Children With Short Stature," published in bile subunit (ALS), have been proposed. Although J Pecf/afr(1995;127:857-867). The first two are ab- Creutzfeldt-Jakob disease is not a complication of stracted in this issue of GGH, however, reading the articles in theirentirety is highly recommended. Letters to the Editors expressing the thoughts and In This Issue opinions of our readers regarding the lead articles of TheSpectrumofUsesforGrowth Hormone this issue of GG/7and related topics are both welcome inChildren page 1 and encouraged. These can be sent c/o Robert M. LetterFromtheEditor page 1 Blizzard, MD, 1224 West Main Street, Suite 701, Char- HowSafeandEffective Is HumanGrowth Hormoneat lottesville, VA22903 orfaxedto 804-977-9450. PhanmacologicDosing? page4 FortheEditoriaiBoard, CME Information page5 RobertM. Blizzard, l\/JD Abstracts page8 Editor-in-Chief This educational activity is supported in part by an unrestricted educational grantfrom Genentech, Inc. . rhGH administration, new safety issues have been aspects of the GH/IGF axis, such as GH-dependent raised, such as tumor recurrence and leukemia. IGF-1 IGFBP-3, andALS. Even an integrated diagno- , Final—ly, new therapeutic options may become avail- sis, however, should be considered provisional, espe- able for example, GH-releasing hormone (GHRH), cially for partial GHD, and the diagnosis must be re- GH-releasing peptide (GHRP), and IGF-1—and the considered in the child who does not respond appro- use of GH must be evaluated in light ofsuch proposed priatelyto conventional GH therapy. It isworth pointing alternativetreatments. out that even though GH has been used in the treatment of GHD for 40 years, there still has never been a APPROVED INDICATIONS FOR GH THERAPY controlled\r\a\ demonstrating the impact of therapy IN CHILDHOOD on adult height. Even when the appropriate diagnosis of GHD has As stated above, there are3 FDA-approved indications been established, controversies remain, such as what for the use of GH therapy to promote growth during the appropriate starting dose of GH should be. While childhood: childhood GHD, growth failure associated studies do demonstrate a dose-response for GH,^ the with CRI, and shortstature associated withIS. Each of slope of this correlation is relatively shallow, with only these diagnosticcategories isworthyofcomment. modest increases in growth rate seen when the GH dosage is increased, for example, from 0.025 to Childhood Growth Hormone Deficiency 0.05 mg/kg/d orfrom 0.05to 0.1 mg/kg/d. On the other hand, it has been argued that there are psychosocial On the surface, this would appearto be the least con- benefits to returning a short child to the normal growth troversial indication for GH treatment. Children with curve as rapidly as possible, and that treatment at the classic GHD have severe growth failure, and GH re- larger dose best assures the eventual attainment of placementshould lead to catch-up growth and the normal adult height.^ Given the significant cost of GH potential for achieving normal adult height. The prob- and the potential for dose-related side effects,^ it lem here lies not in the therapy but in the diagnosis.^ seems best to individualize the therapeutic approach. The potential pitfalls in provocative GH testing have For example, in children who are diagnosed early in been described in detailand are presented in Table 1 life or when short stature is mild, beginning treatment These pitfalls have made the diagnosis of childhood with a dosage of 0.025 mg/kg/d and reserving the op- GHD less clear-cut. tion of increasing the dose if at anytime the growth re- While measurement of serum concentrations of sponse attenuates is most logical. The young child GH-dependent peptides such as IGF-1 IGFBP-3, and with "true" or severe GHD should respond well initially , ALS has certain advantages, currently it is difficult to to the lower dosage (0.025 mg/kg/d). However, forthe GHD GHD determine whether testing these factors is superiorto older child with or the younger child with provocative GH testing in identifying partial GHD and severe short stature, a higher initial dosage of GH GH in predicting the clinical response to treatment.^ (0.05 mg/kg/d) is appropriate. It is important to recog- Biochemical diagnostic strategies are further strained nize, however, that high doses of GH may carry an in- by the testing of children who do not have clear evi- creased risk of side effects or of accelerated entry into dence ofgrowth failure.Accordingly, it is proposed that puberty. In general, it is always best to individualize auxologic criteria be employed as working guidelines the therapeutic approach, with careful assessments of when considering a diagnosis ofGHD (Table 2). clinical responsiveness and potential adverse effects. GHD Thus, the diagnosis of should integrate auxologic and biochemical strategies, which need not be Growth FailureAssociated With Chronic limited to provocative GH testing but can include other Renal Insufficiency In the United States, growth failure associated with Table 1 CRI was the second FDA-approved indication for GH. GH Potential Pitfalls in Provocative Testing This approval was based entirely upon short-term data. While a control group was employed, neither •The nonphysiologicnatureofpharmacologictesting long-term studies nor controlled investigations carried •Difficultyin resolvingconflictingdatafrom2ormoretests out until attainment of adult height have been per- •The inconsistencies in reproducing the response to the formed to date.'' The physiologic basis for growth fail- samepharmacologictests ure secondary to CRI remains unknown, and conse- •Thearbitrarydefinitionofa"normal" response quently, the rationale for GH therapy must be consid- •AgevariabilityandsexsteroideffectoftheGH response ered pharmacologic rather than physiologic. Although •Interassay variations among various GH radioimmuno- several studies have demonstrated increased serum assays concentrations of IGF-inhibitory binding proteins, it is •The impact of nutrition, adiposity, and emotional state on not clear that GH works in these patients by normaliz- theGH response ing IGFBP levels or increasing free IGF-1 Long-term •Costoftests safety issues surrounding the use of GH in children •Risksoftesting,eg, hypoglycemiainduction with CRI and/or following kidney transplantation still need to be addressed. GGH Vol. 14, No. 1—April 1998 2 Short StatureAssociated With Turner Syndrome Table 2 There is little evidence to support the concept of an Auxologic Guidelinesforthe Diagnosis GH endocrine etiology for the characteristic growth failure of Deficiency of patients with TS. Serum concentrations of GH and IGF are normal forage in prepubertal girls with TS, de- •Severe growth retardation (height >3 standard deviations spite the fact that growth failure is typically demonstra- [SD] belowthe mean forchronologic age) inthe absence of ble by midchildhood or earlier.s Considering TS as a analternativeexplanation •Moderate growth retardation (height -2 to -3 SD belowthe form of skeletal dysplasia is probably appropriate, GH mean for age) plus growth deceleration (height velocity and, accordingly, therapy should be recognized as <25th percentile for age) in the absence of an alternative being pharmacologic rather than physiologic. A large explanation numberof studies have demonstrated growth acceler- GH •Severe growth deceleration (heightvelocity <5th percentile arteisopnonwistehdoestnroetamtamtencth itnhaTtSo,bsaelrtvhoeudgihn tnhaeivgerGowHtDh •fAorpargeed)isinptohseianbgsceonncdeiotfioann,aletge,mactriavneiaelxpilrarnaadtiiaotnion, plus children.10 Furthermore, several studies have shown a growth deceleration positive effect of GH on adult height.i^-^s In the •Other evidence of pituitary dysfunction, eg, other pituitary Genentech-sponsored clinical trials, GH-treated deficiencies, neonatal hypoglycemia, microphallus patients had an 8.4-cm increase over their baseline projected adult height; subjects receiving both GH and oxandrolone had a 10.3-cm increase.^3 While this Transition From ChildhoodtoAdult GHD study did not include a placebo control group followed to adult height, a matched historical control group had Treatment of adult GHD is now an FDA-approved indi- a mean adult height identical to that of the treated cation forGH. Whilethe efficacyofGH in correcting the patients' baseline projected adult height. This salutary metabolic consequences of and improving quality-of- effect of GH on adult height has been demonstrated in life issues associated with long-standing adult GHD several other studies,ii 12 although another reported a has been demonstrated, the benefit and/or effect of more modest improvement in adult height. continuing GH treatment in childhood-onset GHD fol- However, most subjects in the latter studiesii-i3 lowing epiphyseal fusion have not been demonstrated. generally have been characterized by a relatively ad- Although it may appear logical to continue therapy vanced chronologic and bone age and/or a relatively without a hiatus, the benefits, if any, of this approach early introduction of estrogen for feminization. Given require analysis through randomized, controlled trials. |the effect of estrogen on epiphyseal fusion, estrogen Unfortunately, significant compliance issues are in- therapy unequivocally will compromise the net growth evitable in an adolescent population towhom no imme- response to GH. From the aspect of maximizing the diately obvious benefit of continued parenteral medica- adult height of each TS patient, the critical issue ap- tion is apparent. Additionally, continuing GH treatment pears to be the number ofestrogen-free years that in adult patients with childhood-onset GHD requires GH is administered. Ideally, therapy should be indi- standardization of the retesting of these patients, since vidualized for each patient in an effort to both normal- as many as 60% to 70% of them will have normal ize growth and adult height and allow pubertal devel- provocative GH resultswhen reevaluated as adults.""s opment at a relatively normal age. This approach requires early diagnosis of TS and initiation of GH treat- THE USE OF GH FOR UNAPPROVED ment by midchildhood or earlier, before the patient's INDICATIONS height falls below the 10th percentile on the normal fe- male growth chart. GH has been used for treatment of short stature associated with skeletal dysplasias, dysmorphic syndromes, metabolic conditions such as hypophos- phatemic rickets, idiopathic intrauterine growth retar- GROWTH, Genetics, & Hormones is published dation (lUGR), and idiopathic short stature (ISS). In the under an educational grant from Genentech, Inc. absence of demonstrable GHD, the growth response The information reflects the views of the editors to GH treatment is generally modest. None of these and/or contributors and not necessarily those of the conditions has been studied with sufficient numbers of sponsor, grantor, orthe publisher. subjects to allow adequate evaluation of the impact of Published by: therapy on adult height; in particular, no concurrent Gardiner-Caldwell control group has been employed in long-term investi- SynerMed gations. This is of particular importance in those conditions forwhich there are insufficient historical datato al- 405 Trimmer Road low comparison of the observed growth response with PO Box 458 the natural history of the disorder and in those disor- Califon, NJ 07830 dersthat are by nature heterogeneous, as in ISS. ©1998Gardiner-CaldwellSynerMed. Allrightsreserved. The issue of GH treatment for ISS is particularly complex.16-19 Undoubtedly, some of these patients 3 GGH Vol. 14, No. 1—April 1998 have constitutional delay of growth and maturation better rationale for dosing should be developed. and can be expected to attain normal adult heights, Therapy for non-GHD short children, such as those even in the absence oftherapy. Furthermore, the limi- with CRI orTS, will continue to entail pharmacologic tations in our ability to accurately diagnose GHD in doses of GH to obtain short-term increases in height. children, as discussed above, may make it difficult to Ultimate adult heights will be increased in TS patients, distinguish between some patients with ISS and oth- but whetherthe same will applyto patients with CRI or ers with partial GHD. There are some children who those with ISS and other causes of short stature have been diagnosed as GHD who are in fact remains to be determined. It is incumbent upon the endocrinologically normal, just as there are some endocrine community, pediatricians, and internists to children categorized as having ISS who have either continue careful monitoring of GH recipients for both partial GHD or IGF deficiency. An additional concern short-term and long-term side effects. As experience with GH treatment of ISS is the issue of whether GH accumulates with GHRH and other GH secreta- therapy results in an earlier onset of puberty than gogues, such as the GHRPs, we will be able to deter- might have othenA/ise occurred, resulting in early epi- mine whether such therapeutic options provide any physeal fusion and the forfeiture of whatever height benefits overGH, even ifonlyto asubsetofpatients. gain might have been attained during the early years of GH treatment.20 These issues make it difficult to REFERENCES strongly recommend GH treatment for patients in 1. RosenfeldRG.JClinEndocrinolMetab1997;82:349-351. these categories. On the other hand, there are un- 2.RosenfeldRG,etal./ClinEndocrinolMetab1995;80:1532- doubtedly ISS or lUGR patients who respond effec- 1540. tively to GH treatment, sometimes in as robust a 3.RosenfeldRG. HormRes1996;46:170-173. manner as GHD patients. Accordingly, it is recom- 4. FrasierSD,etalJClinEndocrinolMetab1981;53:1213-1217. mended that such patients be treated as part of 5.KempSFEndocrinologist1996;6:231-237. prospective clinical trials or on a case-by-case basis, 6. BlethenSL,etalJClinEndocrinolMetab'1996;81A704-1710. following a full discussion ofthe potential benefits and 7. FineRN,etal.fPediatr1994;124:374-382. risks oftherapy. 8. PowellDR,etal.PediatrRes1993;33:136-143. 9. RossJL,etal.JPediatr1985:106:202-206. THE FUTURE OF GH THERAPY 10. RosenfeldRG,etal./Pecfiafr1992;121:49-55. 11. NilssonKO,etalfClinEndocrinolMetab1996;81:635-640. GH It is anticipated that will remain the treatment of 12.HaeuslerG,etal.ActaPaediatr1996;85:1408-1414. ' choice for children with classic GHD, at least for the 13.RosenfeldRG,etal./Pediatrlnpress. nearfuture. Improved formulations, easier methods of 14.VandenBroeckJ,etal.fPediatr1995;127:729-735. reformulation and routes of administration, and long- 15.TauberM,etalJClinEndocrinolMetab1997;82:352-356. lasting GH preparations should all enhance compli- 16.RosenfeldRG./Pediafr1997;130:172-174. ance with and, ultimately, the success of GH treat- 17. LocheS,etal./Pecf/afr1994;125:196-200. ment. As more data are accumulated on the cost:ben- 18. HindmarshPC,etal.Lancet1996;348:13-16. efit ratio of higher GH doses and as more experience 19. HopwoodNJ,etal./PeAafr1993;123:215-222. is gathered on the adverse effects of high GH doses, a 20. KawaiM, etal.fPediatr1997;130:205-209. How Safe and Effective Is Human Growth Hormone Pharmacologic Dosing? at MD Arnold Slyper, lescence.1 Use ofthis dose has led to growth accelera- AssistantProfessorofPediatrics tion in conditions in which GH may be partially defi- MedicalCollegeofWisconsin cient, namely idiopathic short stature,^ as well as con- Milwaukee, Wisconsin ditions in which GH secretion is normal, such as Turnersyndrome,^ bone dystrophies,"* and intrauterine Until 1985, cadaveric growth hormone (GH) was the growth retardation.^ Irrespective of their endogenous sole source of human GH (hGH). The recommended GH status, many short children will experience growth dosage of between 0.24 and 0.3 lU/kg/wk or approxi- acceleration on currently recommended hGH doses. It mately 0.1 mg/kg/wk represented a compromise be- is not surprising, therefore, that conditions other than tween the limited hormone supply and obtaining an classic (severe) GH deficiency (GHD) now account for optimal growth response. Since the introduction of more than 40% of patients treated in this country witl^ " biosynthetic hGH, the recommended dosage has in- biosynthetic hGH. creased to 0.3 mg/kg/wk (0.78 to 0.9 lU/kg/wk). This is It is clearfrom these studies, however, that currently roughly triple that used in the past and about 3 times recommended doses of hGH often have gone beyond the rate of endogenous GH production, except at ado- providing physiologic replacement and are now phar- 6GH Vol. 14, No. 1—April 1998 4 2 macologic. In Turner and other syndromes, for exam- py for several years at the time of diagnosis. Of the 1 ple, GH treatment is clearly aimed at producing a su- cases from Japan, 8 had idiopathic GHD and none of praphysiologic effect, namely, a final height beyond the usual riskfactors for leukemia such as chemother- that dictated by genetic potential.3 This is a unique apy, radiation therapy, or preexisting malignancy. The venture. Never before in the history of medicine has a data from Japan, therefore, remain unexplained.^ It is biologic agent been used in an attempt to produce generally agreed that there is insufficient evidence to such widespread and permanent physical changes. incriminate GH therapy as a cause of leukemia, However, to anticipate that this can be achieved with- leukemic relapse, or tumor recurrence.^ Neverthe- out undesirable side effects also would be unprece- less, a high index of suspicion needs to be main- dented. With this in mind, issues of safety should be of tained, since GH has the potential for being carcino- paramount concern. Unfortunately, there is a limit to genic.7 Acromegalic patients are at increased risk for which the available data from long-term replacement developing benign and malignant tumors, particularly dosing or short-term pharmacologic dosing can be colon polyps and adenocarcinoma.^ In a small group used to guarantee the ultimate safety of pharma- of acromegalic patients with active disease, 53% had cologic GH therapy. colonic polyps.10 In rats, both hypophysectomy and There is a general consensus that over the short large doses of GH influence the effect of carcino- term pharmacologic GH dosing is reasonably safe.^ gensJ Intraperitoneal injection of large doses of puri- An increased risk of intracranial hypertension and fied pituitary-derived GH into rats for up to 485 days slipped capital femoral epiphysis has been docu- resulted in rapid growth as well as neoplasia in multi- mented, but the incidence of these complications is ple organs: lymphosarcomas of the lung, adrenocorti- not large.6 Of all potential complications, however, it is cal and adrenomedullary carcinomas, solid ovarian the risk of malignancy that should be of greatest con- tumors, and breast tumors.'' Such toxicology studies cern. To date, 44 patients have developed leukemia may have little relevance when considering physio- following GH treatment, 12 of them from Japan.6 The logic replacement dosing, but the situation may be malignancies reported have been stem cell malignan- otherwise forpharmacologic dosing. cies, such as leukemias, lymphomas, or thymomas.'' Many of the patients had received pituitary-derived Of possible relevance to this issue are observations GH at moderate doses, and some had been off thera- that melanocytic nevi of children with hypopituitarism and Turner syndrome show increased growth, increased proliferative activity, and atypical signs of dif- CME CERTIFICATION ferentiation during GH therapy, although there is no The GG/yEditorial Board is pleased to announce Cate- evidence of neoplasia. ''2 Increased chromosome gory 1 creditfor GROWTH, Genetics, & HormonesUom fragility also has been demonstrated in lymphocytes the University of Virginia School of Medicine. This en- obtained 3 to 12 months into the treatment of normal during material has been planned and produced In ac- short children, in addition to an increase in sponta- cordancewiththeACCME Essentials. neous chromosome rearrangements and a significant Overview: This enduring material is designed to provide increase in bleomycin-induced aberrations.^3 physicians and other health professionals with current Hyperinsulinemia and insulin resistance have been research and clinical information essential to providing noted in Turner syndrome and normal short children quality patient care to children with growth problems during GH therapy.^^.is The absence of diabetes in all and genetic disorders. but a few patients in no way excludes the possibility of Target Audience: This enduring material is designed for sequelae from a childhood spent in a state of hyperin- pediatricians, pediatric endocrinologists, pediatric ge- sulinemia and insulin resistance.'^ i^ Doubtless, most neticists, and family medicine physicians interested in children will suffer no long-term effects, but this may pediatricgrowth, genetics, and endocrine issues. not be the case for patients who already have a pre- Metliod of Physician Participation: Physicians can disposition to atherosclerosis, diabetes, or hyperten- study each issue of GROWTH, Genetics, & Hormones, We sion. have to admit that our knowledge of the nat- rreeqsupeosntdCtMoEthcerepdoistt-ftoresetascehlfi-sesvuael.uTathieonesqtuiemsattieodnsl,enagntdh ural history of these diseases is limited, and it may be oftimeto completethis enduring material is 1 hour decades before we can say with certainty that treatment has no influence on the development of these Learning Objectives: Through participation in this en- conditions. during materials series, the participant will have the op- Between 53% and 76% of patients with acromegaly portunityto: develop joint problems, with a delay of approximately 1 . Apply current research and advances to the man- 10 years between the onset of acromegaly and the agement of patientcareforoptimum clinical outcomes. appearance of arthropathy.^^ Typical joint changes in- 2. Utilize current research and clinical care issues to clude widening of the joint spaces, osteophyte forma- initiate discussions with colleagues with a focus toward increased awareness of current issues and controver- tion, joint capsule calcification, and mineralization of ligamentous insertions.""^ These changes are irre- sies. versible. Whetherjoint disorganization also occurs in 3. Conceptualize areas forfuture research in thefield ofgrowth and genetics. developing joints as a consequence of higher-dose GH therapy will not be known for years. Reports of 5 GGH Vol. 14, No. 1—April 1998 avascular necrosis of the femoral head and slipped By default, therefore, a subnormal growth velocity capital femoral epiphyses in a few treated children often becomes the decisive factor in the decision to may or may not be the tip of the iceberg.18The poten- initiate GH treatment. However, the measurement of GH tial forjoint disturbancesfollowing high-dose treat- short-term growth velocity is itself subject to biases ment could be a particular concern in conditions with and inaccuracies. Growth velocities in the autumn and preexisting abnormalities of the growth cartilage. winter may be more than 2 cm/y lower than during the Modest short-term growth acceleration has been rest of the year, and a growth velocity of less than achieved in some patients with achondroplasia using 2.5 cm/y during these seasons may be normal.27 One pharmacologic GH dosing.^The basic defect in this study found that growth velocity was significantly GH condition is in the fibroblast growth factor receptor 3 higher after testing than before testing (3.4 cm/y gene and relates in some manner to abnormal carti- versus 5.1 cm/y for prepubertal children and 3.4 cm/y lage growth and endochondral ossification. A bone versus 6.3 cm/y for pubertal children). An explanation dysplasia also contributes to the short stature of for this odd finding may be that growth velocities prior Turnersyndrome, and an abnormalityofcartilage can- to testing were transiently low, leading to a selection not be excluded.19-20GH-treated children with chronic bias in referral.28The 95% confidence limits of a single renal failure could be another group at high risk for height measurement performed by skilled personnel bone and joint complications. The extent to which is ±0.5 cm.29 There is a similar lack of precision for abnormal joints and bones can be increased beyond measuring yearly growth velocities. Fora short normal their genetic potential and yet retain complete func- child growing along the 25th percentile, the confi- tional integritythroughoutadulthood is not known. dence limits for yearly growth velocity span the 8th to Even if we follow a safety-first approach to pharma- 52nd percentiles. In general, the lower limit would be cologic GH treatment, the benefits of therapy are an- considered abnormal while the upper limit would be other important issue. We should not be placing any within the normal range. For measurements taken by child at unnecessarynsk. While higher-dose GH has inexperienced personnel or at less than 12 months improved the height prognosis forchildren with classic apart, confidence limits would be even greater. Over2 (severe) GHD, the situation is far more ambiguous years, there is no correlation between year to year GH with respect to children with other forms of insuffi- growth velocities, suggesting that short-term growth ciency. This is a relevant concern, since these children velocity is an unreliable means of predicting future now constitute a large proportion of children being growth.29 The ambiguities surrounding the diagnosis treated with GH. of neurosecretory dysfunction no doubt account fori NeurosecretoryGH dysfunction wasfirstobserved in some ofthe discrepancies in GH prescribing practices children who had undergone prophylactic cranial irradi- between one pediatric endocrinologistand another. ation for leukemia. Frequent sampling of endogenous Recent interest in treating GHD adults has focused GH over24 hours demonstrated diminished GH secre- attention on the question, "What percentage of pa- tion; GH pulseswereattenuated in size and diminished tients diagnosed with GHD in childhood truly have this in number.2i Similar secretory patterns were subse- condition?"The answershould give pauseforthought. % quently found in other short, poorly growing children Tauber et aPo found that 71 of 98 adults previously who had not undergonecranial irradiation and who had diagnosed as having partial GHD (peak GH response "passed" provocative GH stimulation testing.22 The between 5 to 10 ng/mL) and 36% of 33 adults diag- concept arose of a spectrum of GH insufficiency, rang- nosed with complete GHD (peak GH response ing from short but normallygrowing children atone end <5 ng/mL) had normal stimulated GH peaks of greater of the spectrum, and children with classic GHD at the than 10 ng/mLon a single stimulation test. GHD other, and a group of children with bordertine to sub- Not only is the diagnosis of partial ambiguous, normal growth and partial GHD in between. At the but the results of GH treatment also are unclear. For same time there was a loosening of the "pass-fail" cri- any child receiving GH, puberty appears to be an im- teriaforGH stimulation testing and acutoffof 10 ng/mL portant dividing line in terms of therapeutic response. ratherthan 5to7 ng/mLwasadopted.23 Testosterone and estrogen increase the amplitude of Despite wide acceptance of neurosecretory GH GH pulses, and a pubertal increase in GH accounts in dysfunction as a distinct clinical entity, there is much partforthe growth spurt of puberty.^i This is, however, about this syndrome that is extremely ambiguous. a 2-way relationship, as GH also influences pubertal There are, for example, no objective criteria for its diagnosis. Twenty-four-hour GH monitoring is expen- sive and labor-intensive, and has remained primarily a In Future Issues research tool. It also seems to be no betterat diagnos- MolecularPhysiologyof Leptinand Its Receptor ing GHD than stimulated GH levels.24 The diagnostic YiyingZhong, PhD, and Ron L. Leibel, MD cutoff levels used during GH stimulation testing are Growth HormoneReplacementInAdultGHD Patients recognized as arbitrary.25The finding of substantial PeterSonksen, MD discrepancies between one GH assay and the next, to Insulin, IGF-1 and IDDM: RecentlyImplicated the point that the diagnosis of GHD may depend on Genetic Loci which assay is used, has highlighted the inadequacies Cheryl L. Deal, PhD, andConstantin Polychronakos, MD of provocativetesting.26 6GW Vol. 14, No. 1—April 1998 6 . sex steroid secretion.32 GH treatment of GHD and REFERENCES non-GHD children to final height results in accelerated 1. Kemp SF. Endocrinologist1996;6:231-237 pubertal progression and a pubertal decrease in 2. HopwoodNJ,etal. JPediatr1993;123:212-215. height standard deviation scores (SDS) for bone.^^-^^ 3. NilssonKO,etal. JClinEndocrinolMetab1996;81:635-640. The acceleration in pubertal progression is dose- 4. KeyLL,GrossAJ. Pediatrics1996;128:S14-S27. dependent. In a large group of male children with 5.Albertsson-WiklandK. ActaPaediatrScandSuppl1989; isolated GHD, a doubling of GH dose from 15 lU or 349:35-41. 5 mg/m2/wk to 30 lU or 10 mg/m^/wk increased the 6. BlethenSL,etal. JClinEndocrinolMetab1996;81:1704- rate of pubertal maturation but had no effect on growth 1710. velocity.37 An earlier onset of puberty also was noted 7. SlyperA. MedHypoth 1995;45:523-528. in a controlled study of non-GHD children, and this 8. AllenDB. /Ped/afr1996;128:S8-S13. study concluded that therapy may actually have led to 9.EzzatS,MelmedS. JClinEndocrinolMetab1991;72:245-249. a decrease in final height.34 10.KleinI,etal. AnnInternMed1982;97:27-30. The implication of these observations appears to be 11. BourguignonJ-P,etal. La/iceH993;341:1505-1506. that for children who are not truly GHD, the closer to 12. PierardGE.etal. fPathol1996;180:74-79. puberty that GH treatment is initiated the less likeli- 13.TedeschiB,etal. HumGenet1993;91:459-463. hood of a gain in final height. If treatment is started 14.CaprioS,etal. /Pediatr1992;120:238-243. early in childhood, there is a greater chance of ex- 15.WalkerJ,etal. JClinEndocrinolMetab1989;69:253-258. ceeding genetic potential. However, to accomplish 16.LiebermanSA,etal. EndocrinolMetabClinNorthAm this, supraphysiologic doses of GH need to be admin- 1992;21:615-631. istered throughout childhood, resulting in a greater po- 17. DetenbeckLC,etal. Clin Orthop1973;91:119-127. tential for long-term complications. 18.WatkinsSL. KidneyInt1996;49:S-126-S-127. 19. LippeBM. Primaryovarianfailure. In:KaplanSA,ed. Recommendations ClinicalPediatricEndocrinology.Philadelphia,Pa: Based on this discussion, propose the following rec- WB Saunders/HarcourtBraceJovanvich; 1990:325-326. I ommendations, appreciating that many of these 20.PeltomakiT,etal. JCraniofacGenetDevBiol1989;9:331-338. points are out of line with the current practices of many 21. BlattJ,etal. JPediatr1984;104:182-186. pediatric endocrinologists: 22. SpiHotisBE,etal. JAMA 1984;251:2223-2230. 1. The recommended dosage of GH treatment of 23. RosenfeldRG. JClinEndocrinolMetab1997;82:349-351. 0.3 mg/kg/wk is a high one for the initial treatment of 24. RoseSR,etal. NEnglJMed1988;319:201-207. children with severe (classic) GHD. Treatment should 25.RosenfeldRG,etal. /ClinEndocrinolMetab1995;80: be started at a lower dose and further dose changes 1532-1540. titrated againstthe observed growth effect. 26.ReiterEO,etal. JClinEndocrinolMetab1988;66:68-71. 2. Families of short children who pass provocative 27.MarshallWA. ArchDisChild1971;46:414-420. testing and in whom pharmacologic GH treatment is 28.PolychronakosC,etal. £'ur/Pecf/a^r1988;147:582-583. contemplated should be informed that negative short- 29. VossLD,etal. ArchDisChild1991;66:833-837. term data provide no assurance as to the ultimate 30. TauberM,etal. JClinEndocrinolMetab1997;82:352-356. safety of pharmacologic GH therapy and that the ben- 31.BrookCGD,HindmarshPC. EndocrinolMetabolClinNorth efits of treatment in terms of final height are unknown. Am 1992;21:767-782. Families of children with Turner syndrome who are 32.AlbaneseA, StanhopeR. HormRes1995;44(suppl 3):15-17. about to be placed on the newly recommended GH 33. DarendelilerF,etal. ActaEndocrinol1990;122:414-416. dosage of up to 0.375 mg/kg/wk also should be in- 34. Kawai M.etal. /Pediatr1997;130:205-209. formed that there is little information on the short-term 35. RosenfeldRG. /Ped/atr1997;130:172-173. safetyofthis dose and none on its long-term safety. 36. LocheS,etal. /Ped/atr1994;125:196-200. GH 3. For poorly growing peripubertal children, 37. StanhopeR,etal. HormRes1992;38(suppl 1): 9-13. testing should be accompanied by sex steroid priming This article was revised and updated from an article previously GHD so as to exclude the transient, physiologic pres- published in MedicalHypotheses (1995;45:523-528] with per- ent in many youngsters with constitutional delay of mission oftheeditor. puberty. Sex steroid priming has gone out of favor in recentyears, but could be used far more extensively. Editorial Board 4. A multicenter controllecHr\a\ should be organized Chairman to follow to final height children specifically with neu- RobertM. Blizzard, fAU GH GHD rosecretory dysfunction or partial treated with Charlottesville, Virginia currently recommended doses of GH. It can no longer Associate Editors be taken for granted that these children benefit from William L. Clarke, MD WilliamA. Norton, MD therapy. Noncontrolled studies using estimated heights Charlottesville, Virginia Portland, Oregon or historical controls are incapable of demonstrating JudithG. Hall, MD Fima Lifshitz, MD conclusively the benefits of treatment. In my opinion, a Vancouver, BC, Canada Miami, Florida study of this nature should take priority over other con- AllenW. Root, MD templated growth studies investigating new indications St. Petersburg, Florida GH for treatmentwith ever increasing doses. 7 GGWVol. 14, No. 1—April 1998 Abstracts I'roiii Ihi; Litcn'aturo Short Stature and Growth Hormone Therapy: A National Study ofPhysician Recommendation Patterns The objective of this study was to learn the attitudes of pe- the mean. A normal bone age instead of a delayed bone diatric endocrinologists (PEs) regarding prescribing age increased the recommendation to use GH. Also, boys growth hormone (GH) to short children. Of 534 anony- with comparable shortness to girls, corrected for sex, mous surveys, 434 (81.3%) were returned. Extensive were 1.3-fold more likely to receive GH than girls. planning of the questionnaire permitted the collection and Physicians were sensitive to cost and would have signifi- analysis of data revealing the attitudes of 340 of the 434 cantly increased recommending the use of GH if it cost respondents who currently manage short stature in chil- $2,000 per year instead of $13,000 per year. A cost of dren. Ofthe children currently being treated, 58% were GH $100 per year would have further significantly increased deficient (GHD) and 15% had Turner syndrome (TS). The recommendations. Family wishes clearly influenced the remaining 27% had other causes for short stature. Eight recommendations made by many physicians. In addition, case histories, differing only in physiologic growth vari- the odds of a positive recommendation increased 13% if ables (extent of short stature, growth velocity, normal or the physician believed GH would add at least 1 inch to the abnormal bone age) were presented and the respondents ultimate height of non-GHD children. The authors conclud- were asked whether they were likely to recommend GH for ed from analyses of the data that physiologic factors, each case. Three additional sets of decisions focusing on contingency factors, and belief factors exert independent the contingency variables of price and family wishes also and additive effects on recommendations for GH therapy. were included in the questionnaire. The first 2 contingen- The conclusions that can be drawn from this study have cies proposed that the price of GH therapy fell from ap- several implications for GH and analogous interventions proximately $13,000 per year to $2,000 per year or $100 such as treatment of attention deficit disorder, in vitro fer- per year. In the third contingency, physicians were asked tilization, and genetic testing. Like GH therapy, these anal- their recommendation if the family strongly desired GH ogous interventions hold promise for increasing the therapy, assuming that the price remained at current quality of life fora targeted patient population, but with un- levels. certain risks and benefits and often at considerable finan- Analyses of the data revealed 3 noteworthy patterns in cial cost. As HMOs try to limit the use of GH because of a the responses. First, 68.1% agreed that GH use for non- lack of consensus concerning its use, even referrals for GHD short stature has increased in the past 5 years and short stature may be limited, which would be unfortunate that the physician's knowledge about family finances is since referral not only should address whether GH is rec- marginal in the overall decision-making process whether ommended but also elucidate whether a cause for short to prescribe GH. Second, PEs believe that short stature stature requiring therapies otherthan GH is present. matters and has dysfunctional emotional impact on many children and adults. Third, a lack of consensus existed GuttlerL, etal.JAMA 1996;276:531-537. among the PEs regarding the perceived efficacy (adult height and long-term adverse effects) of GH therapy for Editor's comment: This article demonstrates the optimal non-GHD children. use ofscientific methodology in constructing a question- In applying a logistic model to physicians' decisions to naire that willyield interpretable data. Our readers are en- recommend GH, 3 sets of predictors were used: (a) the couraged to reviewthe entirearticle forits multiple contri- physiologic growth variables previously discussed; (b) butions, particularly the insight ityields into the factors contingency variables, ie, treatment cost and family wish- that mayprompt the prescribing ofotheragents thatmay es; and (c) physicians' beliefs about short stature and GH improve thepatient's qualityoflife butnotnecessarilysig- treatment. The growth rate was very important, as the like- nificantlyimprove the ultimate heightofthepatient. lihood of GH being prescribed increased 3.4-fold for a Although originallypublishedin 1996, thisarticle seemed growth rate belowthe 3rd percentile versus the 3rd to 10th worthy ofabstracting InGGU as itis an excellent corollary percentiles. A height falling below -3 standard deviations to the2leadarticlesin thisissue. (SD) increased by 2.8-fold the likelihood of GH being prescribed than ifthe patient was between -2 and -3 SD below RobertM. Blizzard, MD Considerations Related to the Use ofRecombinant Human Growth Hormone in Children This is an excellent overview of differentaspects ofthe use BACKGROUND of growth hormone (GH) by the Committee on Drugs and Recombinant GH Products the Committee on Bioethics of the American Academy of The biosynthetic process involves a chemical synthesis oil Pediatrics. Some important points are summarized here, the DNA fragment encoding the first 24 amino acids and but the reader is encouraged to review the complete article complementary DNA copies of messenger RNA prepared published in Pecf/a/r/cs 1997;99:122-129. from human pituitarycells. The entire DNAsequence is intro- GGWVol 14, No. 1—April 1998 8 Abstracts From theLiterature duced into a bacterium, Escherichia coli, which enables the side effects, but long-term risks are still unknown. The most synthesis of GH. The products currently available in the frequent side effects are as follows: United States differ in that somatrem (Protropin®, Genen- - antibody formation at an overall level of 10% or lower tech, Inc.) contains an additional methionine group whereas with no clinical effects. somatropin (Nutropin® and Nutropin AQ®, Genentech, Inc., - pseudotumor cerebri related to the use of GH and/or and Humatrope®, Eli Lilly & Co.) are identical to human GH. insulin-like growth factor 1 . Benign intracranial hyperten- Optimal dosing strategies have not been developed fully sion with papilledema has been reported rarely. Cessation of for any product. Most pediatric endocrinologists recom- GH therapy has reversed the symptoms in reported cases. mend 0.18 to 0.30 mg/kg/wk depending on the product, In some cases, spontaneous resolution has occurred in given in equally divided daily doses 6 or 7times aweek. spite of continued treatment. - unusually lean and inappropriately muscular appear- Problems in the Diagnosis of Growth Hormone Deficiency ance due to increased cellular metabolism. Because random fasting serum GH levels do not differ be- - potential physiologic and psychologic trauma related to tween GH-deficient (GHD) and non-GHD patients, other years of regular injections. physiologic and pharmacologic tests have been developed to - theoretical concerns that GH therapy might be related identify GHD patients. Classic severe GHD can be diagnosed to an increased risk of malignancy. Although an increased if the peak stimulated value on 2 tests is < 10 p/L or less in incidence of leukemia among patients treated with GH has association with delayed bone age and slow growth rate. been reported in Japan, a recent US/Canadian survey did Other forms of GHD, ie, partial GHD, cannot be fully diag- not show an increased risk for leukemia or brain tumor un- nosed by these pharmacologic tests. Thus, GH treatment less other risk factors were present, ie, previous radiation should be considered primarily on clinical grounds for those therapy or chemotherapy. patients who present with slow growth velocity and delayed bone age. ETHICAL ISSUES Since there are no data demonstrating improvement in ulti- Goals for rhGH Administration to Children With Short mate height in patients with nonclassic GH, one question Stature should be addressed: Is GH therapy acceptable for children Much of the controversy surrounding the use of GH is who do notfulfill the criteriaforclassic GHD? <f jrelated to the absence of well-defined goals for therapy. It also is well known that being tall is indisputably viewed Most physicians agree that GH therapy should be reserved as a benefit in our culture, and is associated with multiple for patients with either classic GHD or some other condi- advantages, including higher income, academic achieve- tions that exhibit a demonstrable benefit with GH treatment, ment, self-esteem, and social status. The concepts of nor- such as Turner syndrome (TS) or chronic renal insufficiency mality and abnormality are difficult to define, and they in- (CRI). The goal of GH treatment is to attempt to maintain corporate many sociocultural variables. In most cases, a age-appropriate growth and to attain a final adult height that better alternative may be to help children and their families is consistent with the patient's genetic potential. In contrast, to achieve pride and fulfillment on their own terms. other physicians argue that all patients with short stature There also are important economic issues regarding (SS) may be given a trial because of the physical and psy- GH. Some children who do not have classic GHD now chosocial handicaps associated with this condition. receive GH therapy, and these children are likely to come However, there is no universal consensus regarding the def- from the more financially well-off sector of society. On the inition of SS. Some physicians define SS as a height below a other hand, many children lack access to the most basic certain percentile on a standard longitudinal height chart; health care, and it would be ethically inappropriate to spend others believe that height velocity for age and sex is the scarce monetary resources to provide GH therapy to anyone more appropriate indicator. who is notclassically GHD ortruly GH-resistant. Unfortunately, there is an absence of generally accepted criteria for diagnosing "inadequate secretion," partial GHD, CONCLUSIONS AND RECOMMENDATIONS or GH dysregulation. Also, some children with various 1 . Therapywith GH is medicallyand ethicallyacceptablefor: forms of SS respond to GH therapy with accelerated growth -children with classic GHD velocity. Forthese reasons, some pediatric endocrinologists - children with CRI who are awaiting kidney transplanta- believe that GH should be made available for a therapeutic tion trial to all children with SS regardless of its etiology as long -girls with TS as they have a decreased growth rate. - children whose extreme SS keeps them from participat- A survey conducted by the American Academy of Pedi- ing in basic activities of daily living and who have a con- atrics revealed that abnormal GH levels during provocative dition for which the efficacy of GH therapy has been testing followed by decreased growth velocity in SS patients demonstrated. #v were the criteria used most frequently by pediatric endocri- 2. Two key considerations argue against widespread admin- ^nologists. istration of GH therapy to othershort children: -There could be unknown long-term risks. RisksAssociated With GH Therapy -The treatment could result in either no increase or only Short-term treatment with GH has been associated with few an insignificant increase in final adult height. 9 GGH Vol 14. No. 1—April 1998 Abstracts Fromthe Literature 3. Therapy may be justified for children whose height could stimulate parental interest in GH therapy as an avenue for prevent them from participating in the basic activities of improving athletic ability and other forms of social "sue- daily living, cess" fortheirchildren. 4. Pediatricians should be alert to commercial efforts to FimaLifshitz, MD Increased Height in Patients With Medulloblastoma Robertson et al report their surprising findings from chart RobertsonSC, etal. Neurosurge/y1997;41:561-566. reviews of 85 patients with medulloblastomas seen at the University of Iowa College of Medicine from 1963 to 1995. Editor's comment: These authors have presented some These patients (64 children and 21 adults) had their height veryinterestingandintriguingdata. Sincethestudywasret- and weight documented on standardized growth charts be- rospective, there are no carefullycollected hormonal data fore treatment oftheirtumors. The data showthat 22.4% of from the individuals, ie, testosterone, growth hormone, these patients were above the 95% curve (see Table) in IGF-1, etc. Nordo the authors presentany data regarding height. In a comparison group of patients with cerebellar thepubertalstatus ofthechildrenatthe timeofdiagnosis. It astrocytomas, only 7.1% were above the 95% curve for is conceivable thatsome ofthe children with medulloblas- height. Thus, there is a clear difference between linear toma mayhave had earlypuberty which wouldaccount for growth in the 2 different groups. Most of the increased theirbeing taller than expected. Postoperatively growth height was in male patients; however, 56 of the 85 patients failure is the rule ratherthan the exception in these individu- were male. Interestingly, patients who presented as adults als. Although we have no data on finalheights in the chil- also were tallerthan expected at diagnosis. dren, one wouldanticipate that those patients diagnosed The authors related that medulloblastoma cell lines can and treatedas children wouldnotend up being talladults. express different levels of growth factors, including epider- The data do suggest, however thatitis importantforpedi- mal growth factor, platelet-derived growth factor, trans- atricendocrinologists to continue to encourage theirneuro- forming growth factor, and insulin-like growth factor (IGF). surgicalcolleagues to evaluate the hormonalstatus oftheir They note that since the adults in the study also were above patientspreoperative^as wellasaftertreatment. normal height, something must have occurred that pre- dated the development of neoplastic cells. William L Clarke, f^D Preoperative Height and Weight of Patients With Medulloblastomas^ Height Weight Medulloblastoma All Patients PreoperativeCurves (%) M F Total (%) M F Total (%) >95 14 5 19(22.4) 6 0 6(7.1) >90 18 8 26 (30.6) 9 1 10(11.8) >75 36 11 47(55.3) 15 2 17(20.0) >50 46 18 68 (80.0) 32 10 42(49.4) >25 55 25 80 (94.1) 44 14 58(68.2) >10 55 29 84 (98.8) 50 18 68(80.0) >5 55 29 84 (98.8) 54 23 77(90.6) >0 56 29 85(100) 56 29 85(100) ^Numbers in each column representthetotal numberof patientsabove orequal tothe percentile group listed. From Robertson SC, etal. Increased heightin patientswith medulloblastomas. Neurosurgery. 1997;41:561-566. Prenatal Diagnosis From Fetal Cells in Maternal Circulation Detection of fetal aneuploidy by noninvasive means has tein, detect many instances of aneuploidy. However, many been a long-term goal of the prenatal diagnostician. are missed, and this deficit has prompted the search for Screening procedures based on measuring substances in other strategies, including analyzing fetal cells circulating^^ maternal serum, for example, maternal serum (t-fetopro- in maternal serum. Indeed, it has been known for many^P GGH Vol 14, No. 1—April 1998 10